Google Cloud API Gateway Adds OpenAI-Compatible Model Routing in Public Preview — Route to Gemini, Claude, or OSS-GPT Without a Proxy
Google shipped model routing in Google Cloud API Gateway to Public Preview on 2026-08-04: a serverless ingress layer that accepts OpenAI-compatible requests and dynamically routes them to Gemini, Claude, or OpenAI OSS-GPT based on YAML router config, with rate limiting and token tracking built in. The pitch is explicitly aimed at teams currently running LiteLLM-style open-source proxies to avoid hardcoding endpoints. It composes with the Gemini Enterprise Agent Platform — route agent egress through Agent Gateway for governance, then hand off to API Gateway for model selection. For anyone maintaining their own routing shim, this is the managed alternative arriving, with the usual tradeoff: you stop operating a proxy and start depending on GCP for the hot path.
↳ Follow the thread